The highest ROI decision in your global strategy

Full transcript

Welcome and introductions

Marta: Hi, everyone. Thanks for joining today. I'm Marta, Senior Product Marketing Manager here at Lokalise, and I'm super excited to host this webinar on the highest ROI decision in your global strategy — and how you can convince your leadership that you can unlock the highest ROI opportunity in your business.

Today, I'm joined by two excellent speakers: Sophie Krishnan, the CEO of Lokalise, and Sahil Gambhir, CPO at Lokalise. They're going to be speaking about why localization is the highest ROI opportunity and what we've built and brought to market this first half of the year to help you build your business case for your leadership team.

Before we get started: this session is being recorded, and we'll send you the recording via email after the session, so you don't have to worry about that. With all of that said, I'll hide backstage and pass it over to Sophie. I'll be back at the end of the webinar.

Making the business case for localization

Sophie Krishnan: Hello. I'm going to start with a question to understand the journey each of you is on: how confident do you feel making the business case for localization to your leadership team at the moment?

We see a mix of answers, with many people not yet as confident as they'd like — and I'm not surprised. In the last twelve months, I've had quite a few conversations with our customers as they engage with their C-suites about localization in their business. What surfaces most of the time is the following: the benefits of localization seem obvious, the efficiencies from using a modern platform seem obvious, and yet somehow making a case for ROI is not that easy.

We're going to talk today about best practices and about what builds a strong, data-driven case — because today, localizing with the right setup is one of the highest ROI decisions a company operating in several markets can pursue.

I'm Sophie, CEO of Lokalise, and I'll share my perspective today not only as a CEO, but also as someone who has been in many C-suites and on many boards. I'm also lucky to work very closely with our customers on their localization journeys.

How AI changed the big plays

Sophie Krishnan: The C-suite always looks for the big plays: what's going to deliver very high growth with as low effort as possible? It's no surprise to any of you that in the last eighteen months, AI has changed those big plays in many areas — and in localization too, the big play has changed.

Localization is not one of the first areas many C-suites start with. AI typically starts to change the game in areas with high volume and repeatability, like customer support. Then it moves into highly structured areas like engineering, then on to research, insights, and data — and then to localization.

In localization, the destination we all want hasn't changed: how do I create the best experiences for each of my markets, so that customers engage locally with my brand, convert, come back many times, and rave about it? What has changed is how we get there.

Until now, you had to choose the route. You could choose the most scenic route, the fastest, or avoid tolls. In localization, that analogy translates into choosing the route for high quality, or the fastest time to market, or avoiding the effort and the spend — and different customers set their priorities differently across those. For the last twenty years, many have tried to solve that problem: how do we solve for all three at the same time? But until now, you couldn't have high quality, super fast, and low effort to the extent you can today. You had to compromise on some. With AI, the game changed. I'm not saying localization is a pure AI play — I'm saying AI has changed what localization teams can deliver.

Twelve months ago, C-suites started to take a closer look at their business, and localization is increasingly making it to the C-suite agenda. Interestingly, from the conversations I have with our customer champions and their execs, it gets to the agenda with the wrong question: someone saying, "Couldn't we just translate everything through an LLM wrapper?" Instead, what we in this webinar — and many of our customers — are trying to answer is the right question: how can we show the extraordinary ROI that localization done well can generate?

The R in ROI: revenue upside

Sophie Krishnan: Let's talk about the R in ROI — the revenue part. How much upside can we create? Localization has always grown revenue. Twenty years ago, the companies I worked with and their competitors already saw localization as a high revenue generator. It's even more important now, for the same reasons as before — but AI has amplified the necessity.

Start with high returns from discoverability. We already knew that localization drove SEO. It's even more true now, because LLMs optimize for local language and local content. You have to be localized to be discoverable by the LLMs and in AI search results.

Localization also drives higher conversion and loyalty. In a world with a plethora of content, to stand out you need to localize with very high quality content and make your brand truly resonate in your market. There's a lot of good research from places like Nimdzi Insights, which has measured this over the years. A few examples: seven out of ten users always choose their native language when given the option; 71% of users say they don't trust poorly translated content; and 64% of customers would pay more for a product or service in their native language.

At Lokalise, we work very closely with our customers to help them measure the upside from localization, and they attribute between 20% and 50% of revenue growth to localization. That's an average — we even have a segment of about 10% of our customers for whom high quality localization drives more than half of their revenue growth.

The challenge we'll all recognize is that when you improve the localization of the user experience, you're often also improving the product UX and pushing marketing campaigns — so product, marketing, and the localization team all try to claim the upside, and there's no easy way to separate it. Best practice is to do A/B testing or pre-and-post analysis. And again, what we find across our customer base, depending on the setup, is between 20% and 50% of revenue growth.

The I in ROI: the efficiency journey

Sophie Krishnan: Now let's look at the other part of the equation: the investment. I'm going to start with a big watch-out, because we've all read the "90% efficiencies, 97% cost saving" claims. From my experience, you can achieve those kinds of numbers — it's not a myth, and we do have customers who achieve extraordinary efficiencies in their localization. But too many people assume it comes from an OpenAI wrapper or a simple fix. Instead, it's a journey of important, professional steps.

You need the right infrastructure to get the significant efficiencies we're talking about. It's a combination of having the right setup — localization assets, the right platform, workflows, automation — but also high quality AI, meaning AI models that fit your brand, your languages, and your context, plus AI scoring, quality checks, and human-in-the-loop at the right places with the right guidance. It is an infrastructure play.

When we put numbers to this journey — and our research team does an incredible amount of rigorous research with different models, customers, and setups — what we find is: with standard AI and MT with limited context, you get about 40–50% cost reduction. With pro AI with context, smart routing, and human-in-the-loop, you get to about 65% cost reduction. And when you layer on custom AI profiles with advanced RAG, scoring, and human review at the right places, you get to 90–95% cost reduction.

The challenge we all recognize is that it starts with quality. You need to measure quality. The point is not just to reduce cost — it's how we improve the customer experience and the quality of the content for our users while reducing the effort. The reduction in effort has to come with the right quality for your markets.

There's also a side that rarely gets quantified, because it's harder to do: with the right localization platform, you allow teams across the business to significantly reduce clunky collaboration, so you get a lot of efficiencies from time savings and collaboration across the business. And finally, the cost is not always in the budget of those who benefit from the upside — that's actually one of the main challenges today in capturing the right ROI, and I'll come back to it.

Back to the analogy from earlier: we're getting to a world where you can have it all. There are no compromises — you don't have to choose. You can have the most scenic, high quality content route, the fastest time to market, and avoid the tolls and inefficiencies. When we step back, for any of our customers, localization delivers one of the highest ROIs for the business, and it's a no-brainer. There aren't many projects in a business today that will generate 20–50% revenue upside and deliver a 90% reduction in spend and effort. The ROI is very high.

Proving it: one business case, one platform

Sophie Krishnan: The question is: what are the steps we need to take to prove it, and why is it hard? The answer is that there's no silver bullet. There's work to be done, and we have to bring the elements together.

The first element is bringing it into one business case, not isolated ones. Different teams in a business benefit from different parts of the upside: the CRO benefits from higher revenue, the engineering team loves the improved team productivity, the CPO loves the faster time to market, and procurement and the CFO love the reduced LSP costs — but they sit in different parts of the organization. Where our customers have been most successful is when they look at the full ROI across all those teams and bring it as one business case that captures the upside, the efficiencies, and therefore the overall impact on the business.

The second element is bringing it under one platform. Localization is an infrastructure play — not bits and pieces, not a set of wrappers — because quality matters and the right setup matters. When you do it well, you get high quality UX, engagement, higher conversion, and loyalty to your brand. It's a journey for the business, and many of you are already on it. We love helping build that one platform that truly brings the highest impact from localization and drives efficiencies in the business.

As a C-suite executive and board member, I can't help but think that in any business today, localization is the most efficient growth lever you have. It's also become table stakes — and your competitors aren't waiting. We're very keen to engage with any of you; our teams are here to support that journey of driving higher revenue generation and as many efficiencies as possible. Now Sahil, our incredible CPO, will share the solutions you can use to become as successful as possible on that journey and deliver the outcomes your C-suite wants.

Building trust in AI: what Lokalise shipped

Sahil Gambhir: Thank you so much, Sophie. So what have we really brought to the market? We've been shipping with high velocity and, more importantly, with a consistent direction. In winter, we invested in building a world-class AI infrastructure with translations that you can trust. And in spring, we proved its value with clear ROI that you can show to leadership.

Let's look at building trust in AI. The question that's often asked is: can you trust AI with your brand? The straight answer is — not out of the box. As Sophie mentioned, you need the right setup and infrastructure behind it to get high confidence translations, smart orchestration, and results that actually prove it. AI translation is a commodity; it's the combination of multiple underlying capabilities that actually gets you those desired results.

Take the language intelligence layer we've built. It makes translations engine-agnostic: you get the best quality by default, with automatic selection of the top-performing LLM for each content type or language pair. At the same time, you get full flexibility and control — you can manually override routing to use the model that best fits your needs, policies, or legal compliance.

Smart LLM routing matters because the best model for English-to-Italian translation is not the best model for English-to-Arabic translation. Content type matters too: UI copy, marketing copy, and legal disclaimers each respond very differently to different models. Lokalise's dynamic router makes this decision automatically, continuously, and invisibly. The result: industry-first smart routing across five-plus LLM models, delivering 33% higher quality than using ChatGPT alone — based on real acceptance data from thousands of real interactions.

The other side of the language intelligence layer is quality evaluation, or AI scoring. Human review is expensive, but removing it entirely is a quality risk. AI scoring threads that needle automatically: it evaluates every translation against industry-standard MQM criteria, so your team reviews only what the model itself is uncertain about. It protects against overreliance on human review, which is costly, and against blind trust in AI, which can carry quality risks. The result: only 20% of the output requires post-editing, and that portion is flagged. That's up to an 80% reduction in post-editing effort, and review time is cut by nearly half — because the product doesn't only give you a score, it also states why that score is low. We're well aware that while AI quality is improving, we need human-in-the-loop, especially for sensitive content, and to keep costs under control. Our goal is to provide high quality, high confidence translations for our customers.

Generic AI gives you generic output, but your brand isn't generic — it has a tone, an identity, and a life of its own. Custom AI profiles using RAG encode your brand voice, terminology, and past translations into every output. Customers can create multiple profiles, each tailored to the specific needs of their departments and domains — for instance, a separate custom AI profile for legal, one for marketing, and one for software localization itself. These high confidence translations are proven by 90% acceptance rates, which rivals professional human-reviewed translations. These quality results are based on evaluations with enterprise customers using real content and workflows, where human reviewers scored first-pass quality and compared the outcomes of custom AI profiles against their own translations.

And don't take my word for it — here are some quotes from real users, including from compliance-heavy industries like finance and healthcare. One says: "RAG translations are a smart trend because they are sustainable and scalable, providing personalized custom translations. AI profiles and RAG really calmed down a lot of issues we've been having." Another user says: "Overall, the custom AI profile is helping a lot with translations. It's clear that the model is learning from previous examples and getting closer to our preferred style." And a localization lead states: "AI initially gave us an incorrect result. But once manual corrections are made, it accepts the changes and starts reproducing the correct format in similar cases."

Quality AI isn't enough if the process around it is still manual. We now have an MCP server that helps implement Lokalise in your tech stack at lightning-fast speed, removing the barrier of navigating API endpoints and documentation. AI workflows, on the other hand, help automate project management tasks within Lokalise. This ensures there's no tool fragmentation: you get AI-powered localization that runs wherever your team works, and ten-times-faster translation cycles as well.

Spring 2026: proving the ROI

Sahil Gambhir: Let's look at the spring 2026 launches, where we focused on expanding reach and proving results. We shipped automatic quality evaluation: an in-product metrics dashboard that compares custom AI profile results against base Lokalise AI and generic MT, like Google Translate, on your own real data. You see the measurable gap between generic AI and your personalized AI. You can now show leadership the real ROI — not just claim it, but actually demonstrate it. You move from "we believe it works" to "we can confidently show it." It's the most objective proof that custom AI profiles outperform their alternatives.

Again, don't take my word for it, because there's an in-product drill-down that lets you inspect the actual translations behind your evaluation metrics — so you can confidently answer the question of what data these evaluations are based on. This makes the evaluation results more transparent, verifiable, and shareable with internal stakeholders.

What's next: expanding reach

Sahil Gambhir: Let's move on to what's next and our expanding-reach pillar. The right localization solution unlocks potential across your entire organization. Marketing and creative teams are producing more content than ever, and they need a tool that supports their pace and feels natural to adopt. So we now have a purpose-built solution for long-form content — blogs, campaigns, slide decks, internal documentation. You name it, we support it — and there's no specialist bottleneck to adopt and use it. Lokalise is now built for every team, every content type, and every use case, never forcing you to choose between powerful and intuitive.

As we said at the start, every release this year has been a deliberate step in the same consistent direction: localization that works for every team, runs itself, and proves its own value.

Closing remarks

Marta: Thank you both so much. That was super informative, and I'm really excited about what we have coming next. We've reached the end of the webinar. After this session, we'll send you a follow-up email with the recording and the deck. I hope this gives everyone watching a really solid business case to bring to your leadership team, so you can feel confident and really prove that your biggest ROI opportunity is localization. Thank you everyone for coming — thanks, Sahil; thanks, Sophie. I appreciate your time, and see you in the next webinar.